Pseudo-Association Rules algorithm in data mining
Furong Jing, Junhui Yang, XieFu Wen · 2010 3rd International Conference on Biomedical Engineering and Informatics · 2010
In data mining we not only mine the potential association between items, but also estimate the existent association of items which is still reasonable whether or not. Association Rules is mainly used to find the useful knowledge of the items in the mass data. However with the time goes by or the environment changes, previous association rules between data items or initial rules might be not reasonable, so we should check these rules to find the unreasonable ones. An association rule can not find the unreasonable relation between the data items. In this paper, we regard the structure of rules found by association rules as a digraph, and propose the Pseudo-association Rules based on an association rule, which can find the unreasonable rules.